X-ray binaries

Rocket experiments. Sco X-1

Giacconi, Gursky, Hendel 1962

Binaries are important and different!

Wealth of observational manifestations: Visual binaries  orbits, masses Close binaries  effects of mass transfer Binaries with compact stars  X-ray binaries, X-ray transients, cataclysmic variables, binary pulsars, black hole candidates, microquasars… Picture: V.M.Lipunov

Algol paradox

Algol ( β Per) paradox: late-type (lighter) component is at more advanced evolutionary stage than the early-type (heavier) one!

Key to a solution: component mass reversal due to mass transfer at earlier stages!

0.8 M



G5IV 3.7 B8 V

Roche lobes and Lagrange points

Three-dimensional representation of the gravitational potential of a binary star (in a corotating frame) and several cross sections of the equipotential surfaces by the orbital plane. The Roche lobe is shown by the thick line

GS 2000+25 and Nova Oph 1997

On the left On the right – Hα spectrum, – the Doppler image GS 2000+25 Nova Oph 1997 See a review in Harlaftis 2001 (astro-ph/0012513) (Psaltis astro-ph/0410536) There are eclipse mapping, doppler tomography (shown in the figure), and echo tomography (see 0709.3500).

Models for the XRB structure

(astro-ph/0012513)

Evolution of normal stars

Evolutionary tracks of single stars with masses from 0.8 to 150M. The slowest evolution is in the hatched regions (Lejeune T, Schaerer D Astron. Astrophys. 366 538 (2001))

A track for a normal 5 solar mass star

Progenitors and descendants

Descendants of components of close binaries depending on the radius of the star at RLOF. The boundary between progenitors of He and CO-WDs is uncertain by several 0.1M

O .

The boundary between WDs and NSs by ~ 1M O , while for the formation of BHs the lower mass limit may be even by ~ 10M O than indicated.

higher [Postnov, Yungelson 2007]

Mass loss and evolution

Mass loss depends on which stage of evolution the star fills its Roche lobe If star is isentropic (e.g. deep convective envelope - RG stage), mass loss tends to increase R with decreasing M which generally leads to unstable mass transfer.

Evolution of a 5M star in a close binary

Mass loss stages

Different cases for Roche lobe overflow Three cases of mass transfer loss by the primary star (after R.Kippenhahn)

In most important case B mass transfer occurs on thermal time scale: dM/dt~M/ τ KH , τ KH =GM 2 /RL In case A: on nuclear time scale: dM/dt~M/t nuc t nuc ~ 1/M 2

Close binaries with accreting compact objects

LMXBs

Roche lobe overflow.

Very compact systems.

Rapid NS rotation.

Produce mPSRs.

IMXBs

Very rare.

Roche lobe overflow.

Produce LMXBs(?)

HMXBs

Accretion from the stellar wind.

Mainly Be/X-ray.

Wide systems.

Long NS spin periods.

Produce DNS.

Among binaries ~ 40% are close and ~96% are low and intermediate mass ones.

Intermediate mass X-ray binaries

Most of the evolution time systems spend as an X-ray binary occurs after the mass of the donor star has been reduced to <1M O Otherwise, more massive systems experiencing dynamical mass transfer and spiral-in.

The color of the tracks indicates how much time systems spend in a particular rectangular pixel in the diagrams (from short to long: yellow, orange, red, green, blue, magenta, cyan).

(Podsiadlowski et al., ApJ 2002)

IMXBs and LMXBs population synthesis The hatched regions indicate persistent (+45) and transient (-45) X-ray sources, and the enclosing solid histogram gives the sum of these two populations. Overlaid (dotted histogram) on the theoretical period distribution in the figure on the right is the rescaled distribution of 37 measured periods (Liu et al. 2001) among 140 observed LMXBs in the Galactic plane. (Pfahl et al. 2003 ApJ)

Low mass X-ray binaries

NSs as accretors

X-ray pulsars Millisecond X-ray pulsars Bursters Atoll sources Z-type sources

WDs as accretors

Cataclysmic variables • Novae • Dwarf novae • Polars • Intermediate polars Supersoft sources (SSS)

BHs as accretors

X-ray novae Microquasars Massive X-ray binaries

LMXBs with NSs or BHs

The latest large catalogue (Li et al. arXiv: 0707.0544) includes 187 galactic and Magellanic Clouds LMXBs with NSs and BHs as accreting components.

Donors can be WDs, or normal low-mass stars (main sequence or sub-giants).

Many sources are found in globular clusters.

Also there are more and more LMXBs found in more distant galaxies.

In optics the emission is dominated by an accretion disc around a compact object.

Clear classification is based on optical data or on mass function derived from X-ray observations.

If a source is unidentified in optics, but exhibits Type I X-ray bursts, or just has a small (<0.5 days) orbital period, then it can be classified as a LMXB with a NS.

In addition, spectral similarities with known LMXBs can result in classification.

Evolution of low-mass systems

A small part of the evolutionary scenario of close binary systems [Yungelson L R, in Interacting Binaries: Accretion, Evolution, Out-Comes 2005]

Evolution of close binaries

(Postnov, Yungelson 2007)

First evolutionary “scenario” for the formation of X-ray binary pulsar Van den Heuvel, Heise 1972

Common envelope

Problem: How to make close binaries with compact stars (CVs, XRBs)?

Most angular momentum from the system should be lost.

Non-conservative evolution: Common envelope stage (B.Paczynski, 1976)

Dynamical friction is important

Tidal effects on the orbit

(Zahn, 1977) 1. Circularization 2. Synchronization of component’s rotation

Both occur on a much shorter timescale than stellar evolution!

Conservative mass transfer

M=M 1  M 2 

const

Assuming (B.Paczynski):

J orb



M M

1 2

aM



const M

 Change of orbital parameters after mass transfer:

M

1 ' 

M

1   ' 2 

M

2  

M

, 

a a M M

2 

M

1 2    0, if

M

2 >

M

1 0, if

M

2 

M

1

Non-conservative evolution



Massive binaries: stellar wind, supernova explosions, common envelops



Low-massive binaries: common envelops, magnetic stellar winds, gravitational wave emission (CVs, LMXBs)



Stellar captures in dense clusters (LMXBs, millisecond pulsars)

Binaries in globular clusters

Hundreds close XRB and millisecond pulsars are found in globular clusters Formation of close low-mass binaries is favored in dense stellar systems due to various dynamical processes

Isotropic wind mass loss

  Effective for massive early-type stars on main sequence or WR-stars Assuming the wind carrying out specific orbital angular momentum yields: a(M 1 +M 2 )=const  Δa/a=-ΔM/M > 0 The orbit always gets wider!

Supernova explosion

 First SN in a close binary occurs in almost circular orbit  ΔM=M 1 – M c , M c is the mass of compact remnant  Assume SN to be instantaneous and symmetric Energy-momentum conservation 

a f a i

   2 

M

1

M c

 

M

2

M

2    1  If more than half of the total mass is lost, the system becomes unbound

e



M c



M



M

2 BUT: Strong complication and uncertainty: Kick velocities of NS!

Angular momentum loss

•

Magnetic stellar wind.

Effective for main sequence stars with convective envelopes 0.3

Gravitational radiation.

Drives evolution of binaries with P<15 hrs

Especially important for evolution of low-mass close binaries!

Mass loss due to MSW and GW

dL/dt

Axial rotation braking of single G-dwarfs (Skumanich, 1972)

GW~a -4

V

~

t

 1/ 2 Physics : stellar wind plasma ``streams'' along magnetic field lines  2 2  h larger specific angular momentum (Mestel).

Assume the secondary star in a low-mass binary (0.4

 M 2  1.5

M

) experiences m.s.w. Tidal forces tend to keep the star in corotation with orbital revolution:  2   .

Angular momentum conservation then leads to:

dL orb dt



dJ

2

dt

Recolling that:

L orb

 

a

2 and using Kepler's 3d law we get

dt orb

~  2

M

1

a

5 2 

MSW~a -1 MSW is more effective at larger orbital periods, but GW always wins at shorter periods! Moreover, MSW stops when M 2 ~0.3-0.4 M



where star becomes fully convective and dynamo switches off.

P

Binary evolution: Major uncertainties     All uncertainties in stellar evolution (convection treatment, rotation, magnetic fields…) Limitations of the Roche approximation (synchronous rotation, central density concentration, orbital circularity) Non-conservative evolution (stellar winds, common envelope treatment, magnetic braking…) For binaries with NS (and probably BH): effects of supernova asymmetry (natal kicks of compact objects), rotational evolution of magnetized compact stars (WD, NS)

E v o l u i t o n NSs can become very massive during their evolution due to accretion.

Extragalactic binaries

It is possible to study galactic-like binaries up to 20-30 Mpc.

For example, in NGC 4697 80 sources are known thanks to

Chandra

(this is an early type galaxy, so most of the sources are LMXBs).

LMXBs luminosity function

LMXB galactic luminosity function (Grimm et al. 2002) LMXB luminosity function for NGC 1316 (Kim and Fabbiano 2003)

LMXBs luminosity function

Cumulated XLF for 14 early-type galaxies.

(see Fabbianno astro-ph/0511481)

List of reviews

• Catalogue of LMXBs. Li et al. arXiv:0707.0544 • Catalogue of HMXBs. Li et al. arXiv: 0707.0549

• Evolution of binaries. Postnov & Yungelson. astro-ph/0701059 • Extragalactic XRBs. Fabbiano. astro-ph/0511481 • General review on accreting NSs and BHs. Psaltis. astro-ph/0410536 • CVs - Evolution. Ritter. arXiv:0809.1800

- General features. Smith. astro-ph/0701564 • Modeling accretion: Done et al. arXiv:0708.0148

• Population synthesis. Popov & Prokhorov. Physics Uspekhi (2007)